Abstract
While Life’s Essential 8 (LE8) provides a comprehensive measure of cardiovascular health (CVH), its association with mortality among the oldest-old, including centenarians, remains unclear. This study evaluated the relationship between LE8-defined CVH and all-cause mortality across adulthood using data from the China Kadoorie Biobank (Hainan cohort) and the China Hainan Centenarian Cohort Study, including 31,473 individuals aged 30–116. Participants were categorized by life stage and CVH score (low, moderate, high). Higher CVH scores were associated with significantly reduced mortality risk at all life stages, including among centenarians, who experienced a 54.8% lower risk with high CVH. A near-linear dose–response relationship was observed. Population-attributable fractions for mortality reached 36.8% in centenarians. Physical activity and body mass were particularly important in reducing mortality among centenarians. These findings challenge therapeutic nihilism in the oldest-old while underscoring the need for age-specific strategies tailored to distinct physiological profiles is crucial for extending healthy lifespan across the adult life course.
Subject terms: Biomarkers, Cardiology, Diseases, Health care, Medical research, Risk factors
Introduction
Cardiovascular disease (CVD) continues to be the foremost cause of death worldwide1. Extensive epidemiological research has identified a set of modifiable risk factors closely linked to CVD incidence and mortality2–4. In response, the American Heart Association (AHA) introduced the concept of “Life’s Simple 7” to help monitor and improve cardiovascular health (CVH) at both individual and population levels5. This framework was recently updated to “Life’s Essential 8” (LE8), which incorporates eight key components—four health behaviors and four health factors—enabling a more sensitive and comprehensive assessment of CVH6.
A growing body of evidence supports the value of the LE8 score, linking higher CVH scores to lower risks of CVD and other adverse health outcomes7–9. However, most studies confirms that higher LE8 scores predict lower mortality in general and middle-aged populations. Yet, a critical and unanswered question emerges: does CVH still matter at the extremes of human lifespan—specifically, among the oldest-old and even centenarians?
The world’s population is aging rapidly, with the number of adults aged 85 and older expected to double by 2025, reaching 1.2 billion10. The need to understand health determinants in very late life becomes urgent. Although aging increases susceptibility to CVD, it remains unclear whether maintaining better CVH continues to confer survival advantages in the most advanced ages. Understanding the relationship between CVH and mortality across the entire adult lifespan—especially in the very old—is therefore essential for developing effective primordial prevention strategies and supporting healthy aging.
To address this gap, we integrated data from the Hainan cohort of the China Kadoorie Biobank (CKB) and the China Hainan Centenarian Cohort Study (CHCCS), constructing a uniquely wide-aged cohort spanning from middle age to centenarians. This study aims to evaluate the distribution of CVH scores and their association with mortality across different life stages, with particular emphasis on the oldest-old and centenarians. Our findings seek to inform global public health strategies aimed at promoting cardiovascular health and extending healthy longevity.
Results
Baseline characteristics
Table 1 summarizes the characteristics of participants across different life courses. The study sample comprised 31,473 individuals from two cohorts. The CKB study included 20,673 middle-aged adults aged 46.79 (7.46) years and 9007 younger-old adults aged 67.66 (4.48) years. The CHCCS cohort consisted of 798 oldest-old individuals aged 84.99 (4.10) years and 995 centenarians aged 102.76 (2.74) years. Among participants under 80 years of age, those with higher CVH scores were more likely to be younger, female, have higher education (high school and above), and be married (all P < 0.05, Table S3). For participants aged 80 years and older, higher CVH scores were associated with being female and having higher education levels (all P < 0.05, Table S4).
Table 1.
Characteristics of study population stratified by different life courses
| Characteristics | Middle age 30–59 years |
Younger old 60–79 years |
Oldest old 80–99 years |
Centenarian ≥100 years |
|---|---|---|---|---|
| Number (%) | 20673 (69.7) | 9007 (30.3) | 798 (44.5) | 995 (55.5) |
| Age, years | ||||
| Mean (SD) | 46.79 (7.46) | 67.66 (4.48) | 84.99 (4.10) | 102.76 (2.74) |
| Sex (%) | ||||
| Male | 7082 (34.3) | 3711 (41.2) | 320 (40.1) | 178 (17.9) |
| Female | 13591 (65.7) | 5296 (58.8) | 478 (59.9) | 817 (82.1) |
| Education (%) | ||||
| Low | 5119 (24.8) | 5038 (55.9) | 605 (75.8) | 908 (91.3) |
| Middle | 6662 (32.2) | 1710 (19.0) | 119 (14.9) | 67 (6.7) |
| High | 8892 (43.0) | 2259 (25.1) | 74 (9.3) | 20 (2.0) |
| Marital status (%) | ||||
| Married | 19310 (93.4) | 7197 (79.9) | 344 (43.1) | |
| Widowed | 640 (3.1) | 1694 (18.8) | 833 (83.7) | |
| Others | 723 (3.5) | 116 (1.3) | 454 (56.9) | 162 (16.3) |
| Annual income (%) | ||||
| <10,000 yuan | 4403 (21.3) | 2134 (23.7) | ||
| 10,000–19,999 yuan | 7026 (34.0) | 2399 (26.6) | ||
| ≥20,000 yuan | 9244 (44.7) | 4474 (49.7) | ||
| Ethnic group (%) | ||||
| Han | 720 (90.2) | 877 (88.1) | ||
| Minority | 78 (9.8) | 118 (11.9) | ||
| Overall CVH, point | ||||
| Mean (SD) | 74.33 (11.38) | 65.22 (11.93) | 63.47 (12.04) | 56.85 (12.01) |
| Low (0–49) | 594 (2.9) | 966 (10.7) | 107 (13.4) | 295 (29.6) |
| Moderate (50–79) | 12537 (60.6) | 7024 (78.0) | 617 (77.3) | 672 (67.5) |
| High (80–100) | 7542 (36.5) | 1017 (11.3) | 74 (9.3) | 28 (2.8) |
| Diet | ||||
| Mean (SD) | 55.05 (16.02) | 52.96 (16.37) | 53.13 (16.49) | 45.41 (16.22) |
| Low (0–49) | 1846 (8.9) | 1177 (13.1) | 101 (12.7) | 285 (28.6) |
| Moderate (50–79) | 13916 (67.3) | 6002 (66.6) | 534 (66.9) | 620 (62.3) |
| High (80–100) | 4911 (23.8) | 1828 (20.3) | 163 (20.4) | 90 (9.0) |
| Physical activity | ||||
| Mean (SD) | 73.04 (24.75) | 54.06 (23.36) | 53.15 (46.14) | 38.99 (45.71) |
| Low (0–49) | 3608 (17.5) | 3741 (41.5) | 362 (45.4) | 601 (60.4) |
| Moderate (50–79) | 5126 (24.8) | 3123 (34.7) | 56 (7.0) | 54 (5.4) |
| High (80–100) | 11939 (57.8) | 2143 (23.8) | 380 (47.6) | 340 (34.2) |
| Nicotine exposure | ||||
| Mean (SD) | 74.76 (34.55) | 76.71 (32.38) | 85.75 (30.88) | 87.28 (20.44) |
| Low (0–49) | 3416 (16.5) | 1260 (14.0) | 77 (9.6) | 31 (3.1) |
| Moderate (50–79) | 486 (2.4) | 589 (6.5) | 49 (6.1) | 45 (4.5) |
| High (80–100) | 16771 (81.1) | 7158 (79.5) | 672 (84.2) | 919 (92.4) |
| Sleep health | ||||
| Mean (SD) | 83.27 (15.16) | 79.35 (18.54) | 71.54 (24.82) | 71.77 (24.70) |
| Low (0–49) | 1444 (7.0) | 1039 (11.5) | 209 (26.2) | 284 (28.5) |
| Moderate (50–79) | 3004 (14.5) | 1901 (21.1) | 141 (17.7) | 130 (13.1) |
| High (80–100) | 16225 (78.5) | 6067 (67.4) | 448 (56.1) | 581 (58.4) |
| Body mass index | ||||
| Mean (SD) | 72.70 (28.66) | 70.34 (30.67) | 64.07 (42.10) | 40.83 (47.89) |
| Low (0–49) | 1773 (8.6) | 1019 (11.3) | 221 (27.7) | 572 (57.5) |
| Moderate (50–79) | 10372 (50.2) | 4465 (49.6) | 184 (23.1) | 54 (5.4) |
| High (80–100) | 8528 (41.3) | 3523 (39.1) | 393 (49.2) | 369 (37.1) |
| Blood glucose | ||||
| Mean (SD) | 86.30 (29.49) | 74.70 (37.27) | 88.32 (24.00) | 86.68 (23.02) |
| Low (0–49) | 1976 (9.6) | 1896 (21.1) | 93 (11.7) | 100 (10.1) |
| Moderate (50–79) | 2408 (11.6) | 1401 (15.6) | 74 (9.3) | 168 (16.9) |
| High (80–100) | 16289 (78.8) | 5710 (63.4) | 631 (79.1) | 727 (73.1) |
| Blood pressure | ||||
| Mean (SD) | 75.20 (32.21) | 48.41 (38.85) | 28.33 (35.57) | 27.00 (32.64) |
| Low (0–49) | 3171 (15.3) | 3927 (43.6) | 525 (65.8) | 710 (71.4) |
| Moderate (50–79) | 6078 (29.4) | 2842 (31.6) | 192 (24.1) | 204 (20.5) |
| High (80–100) | 11424 (55.3) | 2238 (24.8) | 81 (10.2) | 81 (8.1) |
Education: low, middle and high represent illiterate and elementary, middle, and high school or above, respectively, in CKB, while illiterate, elemental school, and middle school or above in, respectively, CHCCS.
Marital status: others represent separated/divorced/never married in CKB, while widowed, married/divorced/separated/never married in CHCCS.
A clear trend in CVH score across different life courses was observed in Fig. 1. The proportion of participants achieving a high CVH score decreased with age: 36.5% of middle-aged adults, 11.3% of younger-old adults, 9.3% of oldest-old individuals, and 2.8% of centenarians. Conversely, the proportion of participants with a low CVH score increased with age: 2.9%, 10.7%, 13.4%, and 29.6%, respectively. A similar trend was observed for individual CVH metrics, except for nicotine exposure.
Fig. 1.
The percentage of low, moderate and high CVH metric score across life course.
Association assessment
Participants with high or moderate CVH scores had significantly lower cumulative mortality rates compared to those with low CVH scores at different life stages (P < 0.05 for all comparisons, Fig. 2). Restrict cubic splines revealed approximately linear dose-response associations of CVH score with all-cause mortality across the life course (all Pfor non-linear > 0.05, Fig. S2). The risk of all-cause mortality decreased linearly as CVH score increased (all P for linear association < 0.05). Fig. 2.
Fig. 2. Dose-response association between overall cardiovascular health score and all-cause mortality across life course.
In middle age (A), younger old (B), oldest old (C) and centenarian (D). The reference is set at 50 points for the cardiovascular health score, model was adjusted for age, sex, education, marital status, household income/ ethnic group.
Comparing to centenarians with low CVH score, those with moderate score had a 34.7% (HR = 0.653, 95%CI: 0.566–0.753) reduced risk of all-cause mortality, and those with high score had a 54.8% (HR = 0.452, 95%CI: 0.299–0.683) reduced risk of all-cause mortality after adjusting for all potential covariates. The PAFs of high CVH score ( ≥ 80 points) vs. low or moderate score ( < 80 points) was 36.8% in centenarian (Table 2). Similar trends toward reduced risk of all-cause mortality were observed in other age stages. The corresponding risk reduction were 55.9% (HR = 0.441, 95%CI: 0.332–0.586) and 69.2% (HR = 0.308, 95%CI: 0.221–0.431) in middle age; 50.7% (HR = 0.493, 95%CI: 0.433–0.561) and 71.6% (HR = 0.284, 95%CI: 0.220–0.368) in younger old; 42.9% (HR = 0.571, 95%CI: 0.409–0.798) and 44.4% (HR = 0.556, 95%CI: 0.324–0.956) in oldest old. The corresponding PAFs were 23.8% in middle-age, 41.5% in younger old, and 12.0% in oldest old.
Table 2.
Association between CVH score and all-cause mortality across life course
| CVH | N | Deaths | Mortality rate, per 1000 person-years |
Model 1 HR (95% CI) |
Model2 HR (95% CI) |
|---|---|---|---|---|---|
| Middle age, age <60 years | PAF = 23.8% | ||||
| Continue | 20673 | 585 | 2.65 (2.44, 2.87) | 0.969 (0.963, 0.976) | 0.971 (0.964, 0.978) |
| Low | 594 | 57 | 9.18 (7.00, 11.80) | 1.000 (Reference) | 1.000 (Reference) |
| Moderate | 12537 | 407 | 3.03 (2.75, 3.34) | 0.430 (0.324, 0.570) | 0.441 (0.332, 0.586) |
| High | 7542 | 121 | 1.50 (1.25, 1.79) | 0.291 (0.208, 0.406) | 0.308 (0.221, 0.431) |
| P for trend | <0.001 | <0.001 | |||
| Younger old, age 60-79 years | PAF = 41.5% | ||||
| Continue | 9007 | 1432 | 15.60 (14.80, 16.50) | 0.966 (0.962, 0.971) | 0.968 (0.963, 0.972) |
| Low | 966 | 312 | 34.00 (30.40, 37.90) | 1.000 (Reference) | 1.000 (Reference) |
| Moderate | 7024 | 1046 | 14.59 (13.70, 15.50) | 0.481 (0.422, 0.547) | 0.493 (0.433,0.561) |
| High | 1017 | 74 | 6.92 (5.46, 8.61) | 0.270 (0.209, 0.350) | 0.284 (0.220, 0.368) |
| P for trend | <0.001 | <0.001 | |||
| Oldest old, age 80-99 years | PAF = 12.0% | ||||
| Continue | 798 | 225 | 57.04 (49.91, 64.82) | 0.980 (0.970, 0.991) | 0.981 (0.971, 0.992) |
| Low | 107 | 44 | 91.27 (66.89, 20.96) | 1.000 (Reference) | 1.000 (Reference) |
| Moderate | 617 | 162 | 52.33 (44.68, 60.81) | 0.560 (0.401, 0.782) | 0.571 (0.409,0.798) |
| High | 74 | 19 | 51.81 (31.87, 78.72) | 0.545 (0.318, 0.936) | 0.556 (0.324, 0.956) |
| P for trend | 0.005 | 0.012 | |||
| Centenarian, age ≥100 years | PAF = 36.8% | ||||
| Continue | 995 | 923 | 322.62 (302.25, 343.88) | 0.983 (0.978, 0.989) | 0.982 (0.977, 0.988) |
| Low | 295 | 285 | 414.71 (368.41, 464.74) | 1.000 (Reference) | 1.000 (Reference) |
| Moderate | 672 | 613 | 296.96 (274.07, 321.09) | 0.661 (0.573, 0.762) | 0.653 (0.566, 0.753) |
| High | 28 | 25 | 228.34 (150.12, 329.90) | 0.476 (0.315, 0.718) | 0.452 (0.299, 0.683) |
| P for trend | <0.001 | <0.001 | |||
CVH cardiovascular health, CI confidence interval, HR hazard ratio.
Model 1 adjusted for age, sex; model 2 adjusted for education, marital status and household income/ethnic group based on model 1.
Pfor trend: tests for linear trends were performed by modeling CVH category as a continuous variable.
PAF: population-attributable fraction was calculated by the formula PAF=p*(HR-1)/(1+p*(HR-1)), where HR is the hazard ratio comparing CVH score<80 to CVH score≥80 and p refers to the prevalence of CVH score≤80 in the population.
In the analysis of individual CVH score with mortality risk, we observed that age-specific disparity. Among centenarians, higher score of physical activity and BMI were significantly associated with a reduced risk of mortality, the HRs were 0.628 (95%CI: 0.544–0.726) and 0.833. (95%CI: 0.726–0.960) in the high category comparing to low category (Fig. 3). In contrast, among middle age, higher score of nicotine exposure, blood glucose and blood pressure were significantly associated with a reduced risk of mortality, the HRs were 0.670 (95%CI: 0.523–0.858), 0.575 (95%CI: 0.462–0.716) and 0.585. (95%CI: 0.474–0.722) in the high category comparing to low category.
Fig. 3. Association between individual CVH score and all-cause mortality across life course.
In middle age (A), younger old (B), oldest old (C) and centenarian (D). CVH: cardiovascular health; CI: confidence interval; HR: hazard ratio. Model was adjusted for age, sex, education, marital status and household income/ethnic group.
Sensitivity analyses
In the overall populations of the CKB study and the CHCCS, the relationship between higher Cardiovascular Health (CVH) scores and reduced cumulative mortality rates (Fig. S3), reduced risk of all-cause mortality (Table S5), and a linear dose-response association (Fig. S4) remained consistent. This pattern persisted even after excluding participants with a history of CVD (Table S6) or those who died within the first year of follow-up (Table S7) or those who followed up longer 9 years (Table S8). The general associations remained robust when stratified by multiple subgroups, including sex, educational level, marital status, and household income or ethnic group across different life courses. Despite some confidence intervals were not statistically significant due to the limited number of deaths in certain subgroups, the point estimates consistently reflected the main findings (Figs. S5, S6, S7 & S8).
Discussion
Based on the comprehensive age cohort constructed from CKB and CHCCS, we observed the decreasing trend in the proportion of high CVH score and the increasing trend in that of low CVH score across the life course. Adults achieved high or moderate CVH score had a reduced risk of all-cause mortality in different life courses. And elevated CVH scores showed a linear, inverse relationship with all-cause mortality risks. Achieving high CVH score could prevent 12.0-41.5% of all-cause mortality associated with low to moderate CVH score across the life course. Physical activity and BMI play a pivotal role in the mortality risk among oldest old, especially centenarians. This finding fundamentally challenges ageist assumptions and therapeutic nihilism in geriatric care, suggesting that promotive and preventive strategies retain critical importance even at the limits of the human lifespan.
Our analysis demonstrated a clear declining trend in CVH scores across age groups, with mean scores of 74.3, 65.2, 63.4, and 56.9 among middle-aged, younger-old, oldest-old, and centenarians, respectively. This pattern mirrors observations from the NHANES 2013-2018 study, which reported mean scores of 68.7, 62.2, and 62.3 for participants aged 20–39, 40–64, and 65–79 years11. Middle-aged adults show the relatively favorable CVH score compared to that of older adults. The proportion of individuals achieving high CVH was particularly low among the most elderly, with only 9.3% of the oldest-old and 2.8% of centenarians reaching this status, figures lower than those reported in our previous study using LS7 metrics12. The declining trend in CVH scores across age groups may due to the biological aging, multimorbidity, and coexisting geriatric syndromes contribute to the complexity of CVH metric in extremely old populations. Emerging evidence suggests that variability in biological aging and reduced efficiency of homeostatic mechanisms opposing aging may contribute to this clinical heterogeneity. Future studies of gero-therapeutics may identify interventions that can improve CV health, as well as healthy aging and longevity13. Given the suboptimal CVH status of extremely older individuals, our study has significant practical implications for CVD prevention and healthy aging strategies in China, providing basis for estimating healthcare costs and planning healthcare needs in different life courses.
Despite this overall age-related decline in CVH scores, the inverse dose-response relationship with mortality risk remained strikingly consistent. Most notably, centenarians with high CVH scores exhibited a 54.8% lower mortality risk compared to their peers with low CVH. This finding significantly extends existing evidence from studies such as NHANES14, Framingham9, Kailuan7,15, and other researches16–18, which primarily focused on middle-aged and younger-old populations, by demonstrating that the mortality benefit of superior CVH persists into the tenth decade of life and beyond. Notably, this relationship holds particular significant for oldest-old adults and centenarians, who were deemed to survive from the impact of lifestyle factors. Our results suggest that improvements in CVH score, regardless of the life course, can still enhance relative risk profiles for these individuals. The Kailuan cohort study revealed that the benefits of high CVH, gaining longer life expectancy with and without CVD, were persistent across the lifespan from age 45 to 85 years19. A prospective study that enrolled 4499 Chinese adults with mean age at 70 years has demonstrated an inverse association between adherence to ideal levels of CVH score and the risk of all-cause mortality20. Together, these finding underscore the potential for effectively mitigating cardiovascular risk and extending life span through improved overall CVH status, suggesting that achieving and maintaining optimal CVH status is crucial for promoting healthy longevity, even among oldest-old populations. In general, LE8 serve as a comprehensive guide for CVD prevention and health promotion, providing valuable tools for researchers, policymakers, and healthcare systems to monitor and assess CVH at individual and population levels across the life course. Targeted interventions focusing on LE8 factors with lower mean scores are recommended to enhance overall population health.
The population-attributable fractions (PAFs) calculated in our study provide powerful evidence for public health planning. Achieving high CVH scores could prevent 12.0–41.5% of all-cause mortality associated with low to moderate CVH across the life course. The substantial PAF of 36.8% among centenarians is particularly noteworthy, indicating that over one-third of deaths in this population might be preventable through optimization of CVH metrics. Similarly, the high PAF of 41.5% in younger-old adults underscores the importance of early intervention. These findings strongly advocate for integrating LE8-based assessments into national healthy aging strategies and geriatric care models, with tailored approaches for different age groups. The declining trend in CVH scores with age likely reflects the complex interplay of biological aging, multimorbidity, and geriatric syndromes13. Emerging evidence suggests that variability in biological aging and reduced efficiency of homeostatic mechanisms may contribute to this clinical heterogeneity. The persistent protective effect of CVH in extreme aging may operate through multiple pathways, including reduced inflammation, delayed cellular senescence, and enhanced genetic resilience. A recent mediation analysis using UK Biobank data further suggests that ideal CVH can attenuate mortality risk even in individuals with high genetic susceptibility to CVD21, indicating potential interactions at the molecular level that warrant further investigation.
An important finding of this study is the marked heterogeneity in the associations between individual CVH components and mortality risk across different life stages. In middle-aged and younger-old adults, multiple cardiometabolic factors—including nicotine use, hyperglycemia, hypertension, and physical activity—were significantly associated with mortality. This pattern aligns with the well-established evidence during midlife22–24. In striking contrast, among the oldest old and centenarians, only physical activity and BMI remained significant predictors. This attenuation of traditional risk factors in extreme ages may reflect a “healthy survivor” effect, where susceptible individuals have died earlier, leaving a resilient cohort for whom these factors are less prognostic25–29. Additionally, modest elevations in blood pressure or glucose in the very old may be physiologically necessary in the context of frailty30. The persistence of physical activity and BMI as significant predictors of mortality in centenarians warrants particular attention. The U-shaped relationship between BMI and mortality—captured by our LE8 scoring which assigns the lowest category to both BMI ≥ 35 and BMI < 18.5—raises the possibility that the observed BMI-mortality association in centenarians may be driven largely by the hazards of low body weight. Indeed, underweight (BMI < 18.5) in older adults is often a marker of underlying frailty, malnutrition, or occult disease, rather than a reflection of intentional weight loss or healthy leanness31. Similarly, the consistent protective effect of physical activity across all age groups, including centenarians, underscores its fundamental role in maintaining function and delaying mortality, even at the limits of human lifespan. Physical activity preserves muscle mass, cardiovascular function, and metabolic health, and may mitigate the effects of chronic inflammation and frailty32,33. These heterogeneous associations support the development of age-specific strategies that prioritize different components of CVH according to the distinct physiological characteristics and risk profiles of each life stage34. Comprehensive risk factor control remains essential for middle-aged and younger-old adults, whereas for the oldest old, interventions should prioritize maintaining physical activity and preventing unintentional weight loss, while managing overt risk factors on an individualized basis.
The associations demonstrate dose-response patterns across life course, illustrating even modest improvement in CVH score could lead to substantial reductions in mortality risk. Evidence from NHANES revealed the approximately linear dose-responses association of increased total CVH metric score with reduced risk of all-cause and CVD-specific mortality14. Study based on Kailuan also found the inverse does-responses association and identified the optimal threshold of LE8 on death, indicating a significant rise of risk of all-cause mortality to a certain degree7. Benefit accrues at every level of CVH score improvement, with adults at lower initial scores potentially deriving greater benefits from small enhancements. Findings in MESA study demonstrated a linear trend between trajectory groups achieving ideal CVH and the risk of mortality as well17. In essence, those findings emphasize that some improvement in cardiovascular health is invariably preferable to none. Thus, it’s imperative to validate the implementation and assess the utility of this metric in different life stages to enhance public awareness, promote adherence to the recommendations, and ultimately reduce the substantial burden of CVD and mortality.
To our knowledge, this is the first large-scale age cohort study to investigate the association between CVH score based on Life’s Essential 8 and risk of mortality across life course. And the dose-response relationships were observed in different life stages. However, potential limitations should be noted. First, the disparities of two cohorts including the time of baseline surveys and follow-ups, and the methods used for data collecting and determination may limit the generalizability of our findings. Therefore, we conducted corresponding sensitivity analyses and the inherent geographic uniformity, similarity in genetic and environmental backgrounds allow for comparability between age groups. Second, the CKB did not measure blood lips at baseline survey, which limited our analysis to seven metrics when calculating the CVH score. Additionally, the CHCCS had a limited number of deaths from CVD, which restricted our analysis to all-cause mortality. Third, CVH metrics were mainly self-reported and solely measured at baseline, which may introduce recall bias and misclassification bias. Fourth, despite adjusting for multiple variables, residual confounding by measured or unmeasured covariates might have influenced the observed associations. Fifth, the sample size for subgroup analyses within certain CVH score categories was relatively small, which might explain the lack of statistically significant associations. Finally, the generalizability of our findings to the population outside the Hainan province should be approached with caution.
This large-scale cohort study demonstrates that CVH exhibits a strong, inverse association with mortality risk from middle age to centenarians. Critically, even modest improvements in CVH are associated with significant survival benefits, including among the oldest-old. Our findings challenge age-related therapeutic nihilism and reinforce the imperative for targeted health interventions across all life stages. Promoting cardiovascular health through age-specific strategies—comprehensive risk control for younger groups and focusing on physical activity and preventing weight loss for the oldest-old—are essential for extending healthy longevity.
Methods
Study population
The present study utilized two ongoing cohorts in China: Hainan cohort of the CKB study and the CHCCS. Detailed descriptions of the study designs, protocols, procedures and characteristics of both the CKB study and CHCCS have been reported previously35,36. Briefly, the CKB study enrolled 512,891 residents aged 30 to 79 years from 10 regions across China from 2004 to 2008 and a subset of 29,686 participants from Hainan province was selected from the CKB study. The CHCCS, conducted specially in Hainan province, recruited 798 oldest-old subjects aged 80–99 years and 1002 eligible centenarians aged 100 years or older from 2014 to 2017. At baseline, participants in both cohorts were invited to complete comprehensive questionnaires, undergo interviews and physical examinations, as well as provide blood samples. After excluding participants with missing data on cardiovascular metrics, a total of 29,680 adults aged 30–79 years were included in CKB (n = 6 without random blood glucose). Similarly, and 1793 adults aged 80 years or older were included in CHCCS (n = 7 with missing diet, physical activity or nicotine exposure metrics). The detailed flow chart outlining the participant selection process is provided in Fig. S1.
Written informed consent was obtained voluntarily from all participants in both studies. Ethics approvals for the CKB study were obtained from the Ethical Committee of the Chinese Center for Disease Control and Prevention (Beijing, China), the Oxford Tropical Research Ethics Committee, and the University of Oxford (UK). Ethics approvals for the CKB study were obtained from the Biomedical Ethics Committee of Chinese PLA general hospital. These approvals ensured that the research adhered to ethical standards and guidelines for human subject research. All procedures conformed to institutional ethical guidelines and the principles of the Declaration of Helsinki.
Definition of CVH metrics
CVH score and metrics were constructed based on the updated definition provided by the AHA6 and adapted to the methodology of the CKB study and CHCCS.
Dietary information was obtained using a self-reported food frequency questionnaire. In the CKB study, participants who provided complete dietary information for the past 12 months were included in the data analysis. For the CHCCS, participants were required to recall their dietary patterns over the previous 7 days. Based on the Mediterranean Eating Pattern for Americans (MEPA) recommended by the AHA, the criteria were adapted to fit the available dietary variables in the present study. Ten shared diet components were used to construct the diet score. Each component was scored 1 point if the participant met the specified criteria, and 0 point otherwise, resulting in a total dietary score ranging from 0 to 10 (Table S1).
Physical activity (PA) was assessed based on self-reported PA status collected by questionnaire. In the CKB study, weekly time spent on work, transportation, housework, and no sedentary recreation were obtained, and then PA was calculated as metabolic equivalent task hours per day (MET-h/d). In the CHCCS, participants were asked the days of PA in the past 7 days. Mild PA (at least 10-min walk) was used due to the limited proportion of participants engaged in moderate (2.9%) or vigorous (0.6%) PA.
Nicotine exposure was obtained by self-reported smoking history and second-hand exposure from the questionnaire. In both cohorts, the assessment of smoking status compromises current smoker, former smoker and never smoker. Current smoker was defined as smoking at least one cigarette per day or seven cigarettes per week for at least six months. Former smoker was asked years since quitting. Never smoker referred to those who lifetime cumulative smoking amount did not exceed 100 cigarettes/packs.
Sleep health was obtained by self-reported average hours of sleep per night collected by a questionnaire in the both cohorts. In the CHCCS, details of sleep situation over the last month, including bedtime at night, time from going to bed to falling asleep, wake-up time in the morning, and usual actual duration per night, were cross-checked to determine the final sleep health metric score.
Body mass index (BMI) was calculated by the objectively measured weight and height, defined as weight in kilograms divided by height in meters squared. Standing height and body weight were measured with participants wearing light indoor clothing and without footwear.
Blood lipids were excluded for consistency due to only information on taking statins were recorded in the CKB. In the CHCCS, blood samples were utilized to measure the levels of total cholesterol and high-density lipoprotein (HDL) cholesterol.
Blood glucose was measured as fasting blood glucose (FBG) and random blood glucose (RBG) by blood samples. In the CKB study, FBG would be measured the next day when RBG was 7.8 to 11.0 mmol/L. In the CHCCS, only FBG was measured for all participants. Hemoglobin A1c (HbA1c) was calculated based on RBG or FBG using the American Diabetes Association conversion calculator: 28.7×HbA1c (%)-46.7 =Estimated Glucose (mg/dL). Self-reported the diagnosis of diabetes and medication were collected.
Blood pressure was determined by calculating the mean of two measurements taken at rest using an appropriately sized cuff. Self-reported medication of hypotensive drugs was collected.
Quantitative assessment of each CVH metric is shown in Table S2. As recommended by the AHA, the overall CVH score was created by averaging all the individual metric scores to obtained an unweighted score (range, 0–100 points). Then CVH was categorized into 3 groups: scores of 80 to 100, high CVH; 50 to 79, moderate CVH; and 0 to 49 points, low CVH.
Study covariates
In the CKB study, covariates included age, sex (male and female), education (illiterate and elementary, middle school, and high school or above), marital status (married, widowed, and separated/divorced/never married), and household income (annual income <10,000, 10,000–19,999, and ≥20,000 yuan). In the CHCCS, covariates included age, sex (male and female), education (illiterate, primary school, and middle school or above), marital status (widowed, and married/divorced/separated/never married), and ethnic group (Han and minority).
Assessment of outcome
The primary outcome of this study was all-cause mortality. In the CKB study, vital status was obtained through December 31, 2017 determined via official death certificates, which were supplemented with medical records, residential records, active visiting of local communities, hospital records, or verbal autopsies. In the CHCCS, baseline data were linked to mortality data until March 31, 2023 through the death registration system of the public security department and the registration system of the Hainan Provincial Committee on Aging, and verified by telephone.
Statistical analysis
The distribution of baseline characteristics and individual CVH metric scores of the study participants was described across four life courses: middle age (30–59 years), younger old (60–79 years), oldest old (80–99 years) and centenarian ( ≥ 100 years). Additionally, these characteristics were described by the total CVH score, categorized as low (0–49 points), moderate (50–79 points), and high (80–100 points). Descriptive statistics were summarized as percentages for categorical variables and means with standard deviations (SDs) for continuous variables.
The number of follow-up person-years was estimated from the baseline to the end of the study, or until the subject was lost to follow-up or died, whichever occurred first. Mortality rates per 1000 person-years and corresponding 95% CIs were calculated by Poisson regression, stratified by CVH score categories. Kaplan-Meier survival analysis was employed to assess differences in cumulative mortality across CVH score categories, compared using the log-rank test. Cox proportional hazards regression models was conducted to examine the associations between individual and overall CVH score (continuous and categorical, with low level as the reference) and all-cause mortality. Two models were constructed: model 1 adjusted for sex, age, while model 2 adjusted for sex, age, education, marital status, household income (in the CKB study) or ethnic group (in the CHCCS). The proportional hazard assumption was examined using the Schoenfeld residual technique. Adjusted population attributable fractions (PAFs) were estimated for the proportion of all-cause potentially preventable if individuals with low or moderate CVH scores ( < 80 points) attained high CVH scores ( ≥ 80 points). Restricted cubic spline models with three knots were employed to assess the dose-response relationship between the CVH metric score and all-cause mortality, with a CVH score of 50 points as the reference.
Sensitivity analyses were conducted to explore the robustness of the main results and evaluate potential influence from reverse causation. First, all the analyses were repeated in the entire population for each cohort separately. Second, excluded participants with a history of CVD including coronary heart disease and stroke. Third, excluded participants who died within the first year of follow-up. Third, the follow-up time of the CKB was set for 9 years in consistent with the CHCCS. Fourth, subgroup analyses were performed stratified by sex, education level, marital status and household income or ethnic group. All analyses were performed using R version 4.4.1 (R Project for Statistical Computing, https://mirrors.tuna.tsinghua.edu.cn/CRAN/). A two-sided P < 0.05 was considered statistically significant.
Supplementary information
Acknowledgements
The authors sincerely appreciate the contributions of staff and participants at the China Kadoorie Biobank (CKB) and China Hainan Centenarian Cohort Study (CHCCS). This work was supported by the National Nature Science Foundation of China (82173589, 82173590 & 82103920), National Key Research and Development Program of China (2022YFC2503605 & 2023YFC2509400), Key Discipline Talents 4. National Science and Technology Major Project of China (2023ZD0500901), Capital’s Funds for Health Improvement and Research (2022-2G-5031 & 2024-2G-5033), and Beijing Natural Science Foundation (7252181).
Author contributions
Yao He and Miao Liu had full access to all data in this study and take responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Shimin Chen, Miao Liu. Acquisition, analysis, or interpretation of data: Shimin Chen, Shengshu Wang, Bin He. Drafting of the manuscript: Shimin Chen, Shengshu Wang, Bin He, Shanshan Yang. Critical revision of the manuscript for important intellectual content: Ke Han, Yueting Shi, Huaihao Li, Yinghui Bao, Junhan Yang, Miao Liu, Yao He. Statistical analysis: Shimin Chen, Wenchang Wang, Shengyan Du, Xiaoxin Ye, Yanhao Wan, Jianhua Wang, Yali Zhao, Yizhi Chen, Qiao Zhu, Chaoxue Ning, Yujian Chen, Shihui Fu, Xiangyang Zheng. Obtained funding: Miao Liu, Yao He. Administrative, technical, or material support: Jianhua Wang, Dianjianyi Sun. Supervision: Miao Liu, Yao He.
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due to privacy or ethical restrictions but are available from the corresponding author on reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Miao Liu, Email: liumiaolmbxb@163.com.
Yao He, Email: yhe301@sina.com.
Supplementary information
The online version contains supplementary material available at https://doi.org/10.1038/s41514-026-00395-5.
References
- 1.Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990-2021 A systematic analysis for the Global Burden of Disease Study 2021. Lancet.403, 2100–2132 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Li, S. et al. Modifiable risk factors associated with cardiovascular disease and mortality in China: a PURE substudy. Eur. Heart J.43, 2852–2863 (2022). [DOI] [PubMed] [Google Scholar]
- 3.Joseph, P. et al. Cardiovascular disease, mortality, and their associations with modifiable risk factors in a multi-national South Asia cohort: a PURE substudy. Eur. Heart J.43, 2831–2840 (2022). [DOI] [PubMed] [Google Scholar]
- 4.Lopez-Jaramillo, P. et al. Risk factors, cardiovascular disease, and mortality in South America: a PURE substudy. Eur. Heart J.43, 2841–2851 (2022). [DOI] [PubMed] [Google Scholar]
- 5.Lloyd-Jones, D. M. et al. Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association’s strategic Impact Goal through 2020 and beyond. Circulation121, 586–613 (2010). [DOI] [PubMed] [Google Scholar]
- 6.Lloyd-Jones, D. M. et al. Life’s Essential 8: updating and enhancing the American Heart Association’s Construct of Cardiovascular Health: a presidential advisory from the American Heart Association. Circulation146, e18–e43 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jiang, J. et al. Association of Life’s Essential 8 with all-cause mortality and risk of cancer: a prospective cohort study. BMC Public Health24, 1406 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mahemuti, N. et al. Urinary albumin-to-creatinine ratio in normal range, cardiovascular health, and all-cause mortality. JAMA Netw. Open6, e2348333 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rempakos, A., Prescott, B., Mitchell, G. F., Vasan, R. S. & Xanthakis, V. Association of life’s essential 8 with cardiovascular disease and mortality: The Framingham Heart Study. J. Am. Heart Assoc.12, e030764 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.(WHO), W. H. O. Ageing and Health, https://www.who.int/news-room/fact-sheets/detail/ageing-and-health (2022).
- 11.Lloyd-Jones, D. M. et al. Status of cardiovascular health in US adults and children using the American Heart Association’s New “Life’s Essential 8” metrics: prevalence estimates from the National Health and Nutrition Examination Survey (NHANES), 2013 through 2018. Circulation146, 822–835 (2022). [DOI] [PubMed] [Google Scholar]
- 12.Liu, M. et al. Ideal cardiovascular health in the oldest-old and centenarians and its association with disability and health-related quality of life. Front Cardiovasc. Med.8, 603877 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kumar, M. et al. Life’s essential 8: optimizing health in older adults. JACC Adv.2, 10.1016/j.jacadv.2023.100560 (2023). [DOI] [PMC free article] [PubMed]
- 14.Sun, J. et al. Association of the American Heart Association’s new “Life’s Essential 8” with all-cause and cardiovascular disease-specific mortality: prospective cohort study. BMC Med.21, 116 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xing, A. et al. Life’s Essential 8’ cardiovascular health with premature cardiovascular disease and all-cause mortality in young adults: the Kailuan prospective cohort study. Eur. J. Prev. Cardiol.30, 593–600 (2023). [DOI] [PubMed] [Google Scholar]
- 16.Hernández-Martínez, A. et al. Cardiovascular health in Spain based on the Life’s Essential 8 and its association with all-cause and cardiovascular mortality: the ENRICA cohort. Rev. Esp. Cardiol. (Engl. Ed.)77, 372–380 (2024). [DOI] [PubMed] [Google Scholar]
- 17.Sheng, Q. et al. Cardiovascular health trajectories and subsequent cardiovascular disease and mortality: the multi-ethnic study of atherosclerosis (MESA). Am. J. Prev. Cardiol.13, 100448 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Isiozor, N. M., Kunutsor, S. K., Voutilainen, A. & Laukkanen, J. A. Life’s Essential 8 and the risk of cardiovascular disease death and all-cause mortality in Finnish men. Eur. J. Prev. Cardiol.30, 658–667 (2023). [DOI] [PubMed] [Google Scholar]
- 19.Xia, X. et al. Cardiovascular health and life expectancy with and without cardiovascular disease in the middle-aged and elderly Chinese population. BMC Public Health23, 2528 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chen, S. et al. Association of cardiovascular health at old age with all-cause mortality: a prospective cohort study in China. BMC Geriatr.23, 437 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhang, J. et al. Relation of Life’s Essential 8 to the genetic predisposition for cardiovascular outcomes and all-cause mortality: results from a national prospective cohort. Eur. J. Prev. Cardiol.30, 1676–1685 (2023). [DOI] [PubMed] [Google Scholar]
- 22.Le Foll, B. et al. Tobacco and nicotine use. Nat. Rev. Dis. Prim.8, 19 (2022). [DOI] [PubMed] [Google Scholar]
- 23.Qiu, Z. et al. Phenotypic heterogeneity of type 2 diabetes and risks of all-cause and cause-specific mortality. Cell Rep. Med.6, 102367 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Martinez-Gomez, D. et al. Physical activity and all-cause mortality by age in 4 multinational megacohorts. JAMA Netw. Open7, e2446802 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fall, T. et al. Age- and sex-specific causal effects of adiposity on cardiovascular risk factors. Diabetes64, 1841–1852 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wan, E. Y. F. et al. Age-specific associations between systolic blood pressure and cardiovascular disease: a 10-year diabetes mellitus cohort study. J. Am. Heart Assoc.9, e015771 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wang, T. et al. Age-specific modifiable risk factor profiles for cardiovascular disease and all-cause mortality: a nationwide, population-based, prospective cohort study. Lancet Reg. Health West Pac.17, 100277 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Tromp, J. et al. Age dependent associations of risk factors with heart failure: pooled population based cohort study. Bmj372, n461 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Burgess, S. & Zhou, A. Genetic predictors of traits in elderly subjects: risk of survival bias and reverse causation. Eur. Heart J.45, 2155–2157 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chia, C. W., Egan, J. M. & Ferrucci, L. Age-related changes in glucose metabolism, hyperglycemia, and cardiovascular risk. Circ. Res.123, 886–904 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Gao, Y. et al. The relationship between frailty, BMI, and mortality in older adults: results from the CLHLS. BMC Geriatr.25, 539 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Thyfault, J. P. & Bergouignan, A. Exercise and metabolic health: beyond skeletal muscle. Diabetologia63, 1464–1474 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Valenzuela, P. L. et al. Exercise benefits in cardiovascular diseases: from mechanisms to clinical implementation. Eur. Heart J.44, 1874–1889 (2023). [DOI] [PubMed] [Google Scholar]
- 34.Tian, F. et al. Ranking age-specific modifiable risk factors for cardiovascular disease and mortality: evidence from a population-based longitudinal study. EClinicalMedicine64, 102230 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Chen, Z. et al. China Kadoorie Biobank of 0.5 million people: survey methods, baseline characteristics and long-term follow-up. Int. J. Epidemiol.40, 1652–1666 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.He, Y. et al. Cohort profile: the China Hainan centenarian cohort study (CHCCS). Int. J. Epidemiol.47, 694–695h (2018). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to privacy or ethical restrictions but are available from the corresponding author on reasonable request.